Knowledge based support systems, to assist crop management decisions, are required to ensure that the treatments applied are economically advantageous and environmentally acceptable. The knowledge base may contain predictive models for various aspects of crop management such as crop growth, disease susceptibility, nutrient and irrigation requirements. Agrometerological data is an essential input to the knowledge base and there is a need to develop a standardized format for this that will allow interfacing with predictive models and farm management systems. Work is being done to select and integrate some well tested predictive models with agrometerological data networks to create interactive decision support systems for farm management.
Software is being developed for 3 crops, namely, wine graphs, apples and wheat that can be used to improve the accuracy of farmers' weather related decisions, in particular, the timing and frequency of chemical applications and irrigation. Formats of meterological data used in agrometerological networks have been collected and compared. A programme is being developed for transferring data from weather stations to microcomputers via telephone lines for storing the data in a structural query language (SQL) database. Predictive models have been surveyed for wine graphes, apples, wheat and irrigation and some of these are being adapted for implementation. These include models for evapotranspiration, water balance and plant development and models covering a range of vine diseases. In particular, the model of risk analysis for plasmopara viticola is being developed for the north of Italy.
At present, wide scale implementation of P.M.D.A. is hindered by various causes. For example insufficient established and adapted models, insufficient integration of previsional models with agrometeorological data networks and insufficient standardization of interfaces to agro meteo networks. This situation calls for a well thought out intervention to coordinate model testing and implementation.
The tasks of project SYBIL are :
a) attain integration of well tested Previsional Models for wheat, wine, apple and models for irrigation into regional Agro-meteorological data networks within interactive Decision support systems for individual farm adapted crop management advice to fill the gap between models and networks;
b) establish a methodology for ensuring modularity, portability and extensibility to additional regions of P.M.D.A. by introducing uniform software engineering, including uniform choices for data structures, computer platform and user interface;
c) to disseminate validated P.M.D.A. addressing preeminent application - irrigation, fertilization, pest control - making them available for field testing and use by technicians in export regions with regional model adaption.